Case Study: Spotify Ads Ways of Working

Strategy, Process, Storytelling

This project focused on redefining the product lifecycle for the Spotify Ads product org, including the behaviors, rituals, and artifacts from the XFN product teams that support it.

I was given the opportunity to lead our quartet - that’s PM, Eng, Design, and Data Science- to work more efficiently and produce higher-quality products.

The approach I took was to first align our leadership layer — my peers and all of the leaders we supported a level down — on a shift away from thinking of alignment and autonomy as two different ends of a spectrum. As in, more autonomy equals less alignment.

Instead, we have to think of alignment and autonomy as two different dimensions.

In low-alignment, low-autonomy cultures (bottom left), we have micro-management scenarios. Our people aren’t sure why it is they’re doing what they’ve been told to do. There’s no higher-level purpose, and squads at the ground level feel undervalued.

With High alignment, and low autonomy, (at the top-left) the team understands the challenge, but the solution is prescriptive and lacks any space for creative problem-solving. And that might mean you’re not innovating.

In high autonomy and low alignment cultures, bottom right, squads do whatever they want and often run in different directions, building in silos and creating technical and design debt with disconnected work tracks. Leaders are helpless here and spend much of their time helping their teams piecemeal the product together.

So at the top-right is what we needed to become: High alignment, and high autonomy. In this culture, the leader’s job is to communicate what problem needs to be solved and support their teams to understand why. From there, they unblock their teams who are now more likely to collaborate with each other to find the right solution.

So with the meta alignment on the culture we needed to build, I led the charge in redesigning our process. The good news was that I didn’t have to start from scratch. After connecting with other leaders from across Spotify, I learned that we already had a great shared process called The Scale. Now, being that it's a process aimed at solving the design phase, it lacked the specifics needed for a healthy end-to-end process. Such as goals for each phase and inputs from each discipline throughout. Nor did it have stated expectations for behaviors and artifacts we as leaders need to see, and how we measure success.

So that was my initial focus. I partnered with my PM and ENG counterparts, plus our operational leader who would be responsible for rolling it out, and we set up working sprints over 3 months to uncover our gaps and define the goals and expectations for each phase—turning this from Spotify’s design process into our BU’s product development process.

Below are the high-level goals for each phase, and some additional breakdowns.

The other big tweak to our product lifecycle was introducing phase gates. Call these product reviews if you will, but we needed to call them “gates” to deter bad actors from doing what they thought was best, and instead do what’s best for achieving our business goals. And while it might feel like these gates would slow teams, we quickly learned that all these alignment nudges meant more agreement which meant teams moved faster.

So what exactly does all of this work yield? In our first formal review of the process, these were the key wins we highlighted during our annual all-hands: